Energy Management-Oriented Engine Thermal Model in Hybrid Powertrains

抄録

An Internal Combustion Engine (ICE) experiences higher injection quantity and extra friction loss during low operation temperatures. This feature provides further potential for improving the engine fuel efficiency with the additional control freedom in hybrid powertrain configurations. For energy management purposes, this paper presents a state-space engine thermal model that reflects the effects of low temperature. The system states include the engine coolant temperature, which reflects the excessive gas injection, and the lubricant temperature that fluences the extra friction loss. The thermal dynamical system's uncertain parameters are calibrated using the homotopy Levenberg–Marquardt algorithm for minimized residual RMS through the bench tests under pre-defined engine operation points. Their relationships with the operation point are regressed with linear functions. Exponential functions are used to describe the relationship between excessive injection/friction and thermal states. Validation with repeated tests shows that the established model can reasonably predict the engine's thermal behavior. Finally, we integrate the engine thermal model into a hybrid vehicle simulation model. Simulation results under a randomly generated driving circle reveal about 20% more fuel consumption considering the cold temperature influences.